research report - wordpress.com€¦ · cognitive brain research 6 1998 321–334 . research report...

Ž .Cognitive Brain Research 6 1998 321–334

Research report

Electrophysiological evidence of a perceptual precedence of global vs.local visual information

Alice Mado Proverbio a,), Alessia Minniti a, Alberto Zani b

a CognitiÕe Electrophysiology Laboratory, Department of Psychology, UniÕersity of Trieste, Via dell’UniÕersita’ 7, 34123 Trieste, Italyb CognitiÕe Electrophysiology Laboratory, Istituto di Psicologia, Consiglio Nazionale delle Ricerche, Viale Marx, 00137 Rome, Italy

Accepted 9 December 1997

Abstract

Aim of the present study was to investigate the mechanisms of attentional selection of hierarchically organized visual patternsŽ .compound letter stimuli , while subjects were engaged in target selection at either the global or local level. Event-related brain potentialsŽ . Ž .ERPs were recorded using a high density electrode montage. Reaction times RTs to target stimuli were also recorded. RT dataindicated the interference effect of global incongruent information with the local one. ERP data were consistent with behavioral data. Infact, the early sensory N115 component recorded at the primary visual areas exhibited smaller responses to locally attended elementswhen the global configuration was incongruent rather than congruent, suggesting an interference effect of the global with the local level.Conversely, no interference effect was found for globally attended configurations. These results strongly support the view of a perceptualadvantage of globally conveyed information, very likely mediated by low spatial frequency channels. At later processing levels, N1 andP3 components were faster and larger when attention was paid to the global configuration. The difference between target and nontargetresponses, indexing the attentional target selection, yielded a broad occipital–temporal negativity focused onto the left hemisphere in theattend-local, and over the right hemisphere in the attend-global condition. The present findings indicate a hemispheric asymmetry incerebral activation during localrglobal processing. In addition, they provide robust evidence of a sensory precedence of globalinformation. q 1998 Elsevier Science B.V.

Keywords: Compound stimuli; Global and local visual information; Attentional selection; ERPs; N115; Selection negativity; Global advantage;Hemispheric asymmetry

1. Introduction

An interesting issue in current literature concerns howthe brain generates visual unitary percepts. The questionhas been arisen whether different perceptual mechanisms

Žare involved in the construction of a unitary figure for.example, a ‘keyboard’ by organizing local elements as

well as in the segmentation of different elements of aŽ . Ž .configuration a single ‘key’ . Using reaction time RT

w xmeasures, Navon 16 showed that visual information isprocessed in a top–down order. The analysis of the globalconfiguration had an advantage on the analysis of localelements. The author adopted hierarchically structured

Ž .stimuli, consisting in large global configurations, eitherŽ .geometrical or linguistic, composed of smaller local ele-

ments to allow the selection of either the global or local

) Corresponding author. Fax: q 39-40-312272; E-mail:[email protected]

level of information. Typically, when a subject is asked toidentify a target stimulus at either the global or local level,the results are that global patterns are identified faster thanlocal ones, and that when the global configuration isincongruent with the local element it interferes with theidentification of the latter, leading to slower responsetimes. These data have been interpreted as an index of ageneral advantage of global vs. local visual information.Some neuropsychological and electrophysiological studieshave questioned that the global advantage and the interfer-ence effects are expression of the same phenomenon: theperceptual advantage of the global information. For exam-

w xple, Lamb and Robertson 13 in a divided attention taskcarried out with normal subjects found a RT advantage forlocal stimuli subtending large visual angles, along with aglobal interference effect independent of visual angle. Thispattern of results made them conclude that the RT advan-tage and interference are dissociable effects, and do notnecessarily reflect a perceptual precedence of global vs.

0926-6410r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved.Ž .PII S0926-6410 97 00039-6

( )A.M. ProÕerbio et al.rCognitiÕe Brain Research 6 1998 321–334322

local information. Rather, it would be founded on post-per-ceptual attentional mechanisms. More recently, the psy-

w xchophysical study by Hughes et al. 9 have providedstrong evidence that global precedence may actually beperceptual in nature, it being conveyed by low spatialfrequency channels. According to the authors, low spatialfrequency content would be available before informationcarried by high frequency bands, also because of the highcontrast gain of the magnocellular pathway. This wouldexplain why processing of the global configuration isfaster than that of the local elements, and is not greatlyaffected by conflicting local cues.

The nature of global advantage, whether perceptual orattentive, is an object of great debate in the literature. Therole of sensory mechanisms in the global advantage, and inparticular that of a different sensitivity and transmissiontime of spatial frequency channels, has been demonstrated

w xfor instance by Lagasse 12 using high-pass filtered stim-w xuli, and by Hughes et al. 8 using contrast-balanced dots.

Other studies have proposed that an asymmetric inhibitionbetween spatial frequency channels might be potentiallyresponsible for a sensory advantage of low frequency

w xpatterns, as for instance advanced first by Hughes 7 andw xlater on by Lovegrove et al. 15 . According to the atten-

tional hypothesis, however, the global advantage would bemediated by post-perceptual mechanisms, and global andlocal information would be encoded in parallel and avail-

w x w xable at the same time 3,4 . Rather recently, Hubner 6¨provided evidence supporting the intriguing hypothesisthat both sensory and attentional mechanisms might beresponsible for the global advantage. In his study, theauthor adopted both unfiltered and high-pass filtered com-pound stimuli. He found that low frequency content wassufficient but not necessary to elicit a global advantage,while the attentional set, manipulated by keeping eitherrandomized or fixed the target level in different experi-ments, produced a strong interaction between filtering andglobalrlocal processing.

Goal of the present study was to investigate how selec-tive attention modulates the perception of hierarchicalstimuli by means of electrophysiological and behaviouraltechniques. In particular, the experiment was designed totest the aforementioned models of perceptual vs. atten-tional precedence of globalrlocal analysis. Previous elec-

w xtrophysiological studies 1,3,4,10 have indicated a poste-rior N250 component as the first sign of independentprocessing of the globalrlocal levels. In the experimental

Ž .paradigm usually adopted i.e., a divided attention task ,subjects were asked to respond to a target letter indepen-

Ž .dent of its hierarchical level global or local , so that bothlevels were always targets. Thus, the effect of interferencewas evaluated indirectly by comparing the interferencegiven by the introduction of stimuli and distractors charac-terized by a different degree of perceptual similarity. In the

w xstudy by Deruelle and Neville 1 , the onset of the lateposterior negativity related to target selection was earlier

in the global than local condition, while Johannes et al.w x w x w x10 , Heinze and Munte 3 and Heinze et al. 4 found no¨clear pattern of either advantage or interference betweenthe two levels. The latter studies only provided evidence ofseparate mechanisms for globalrlocal levels at late stagesof information processing.

In their electrophysiological study, Ridderinkhof andw xvan der Molen 19 adopted a choice reaction time task in

which the target level was kept fixed within a given run,and subject had to discriminate between two targets. Theauthors considered only the latency of late P300 compo-nents, and contrasted them with the onset of readiness

Ž .potential and response-locked electromyogram EMG toassess whether the locus of interference of the irrelevantlevel was perceptual or induced by response competition.They failed to provide direct evidence of a global advan-tage on the local level. The results also showed no asym-metry in the magnitude of global vs. local interferencebecause of the large intra-subjects variability. Interestingly,they indicated no effect of stimulus incongruence on thelatency of readiness potential, thus, suggesting a perceptuallocus for the interference effect.

In the present experiment, we adopted a selective atten-tion task in which subjects were told to respond to targetletters at a given level, either the global or local one, andto ignore the identity of the constituent letters in theopposite level. This paradigm allowed us to measure di-rectly the effect of interference given by the perceptualincongruence of the unattended level with the attended

Ž .one. Event-related potentials ERPs of the brain to rele-vant and irrelevant stimuli, as well as RTs to targets, wererecorded to monitor the time course and topographicaldistribution of brain activity during processing of com-

Žpound stimuli as a function of the attended level local vs..global or incongruence. We expected that if the global

advantage was perceptual in nature, there had to be anŽ .effect of attended level and incongruence or both on the

sensory responses of visual areas earlier than N250.

2. Materials and methods

2.1. Subjects

Eight right-handed volunteers served as subjects in thisexperiment. All had normal vision or a mild myopiacorrected by glasses. Their ages ranged among 24 and 33years. One subject was excluded from statistical analysesfor exceeding ocular artefacts.

2.2. Stimuli

X Ž .Stimuli were 38=28 15 large capital letters H or SŽ . X Xmade up of smaller letters H or S subtending 22 =15 of

visual angle. Small letters were arranged in a 5=5 matrixw xas in Lamb’s and Robertson’s study 13 . Because letters H

Ž .and S at the global level i.e., large letters were combined

( )A.M. ProÕerbio et al.rCognitiÕe Brain Research 6 1998 321–334 323

with H and S letters at the local level, there were fourŽ .stimuli in all: HrH global H and local H , HrS, SrS,

SrH. Stimuli were white on a black background; theirluminance, measured on small letters, was 2.4 cdrm2.They were randomly presented for 100 ms at two eccentric

Ž .locations i.e., 28 from their external border of the rightand left visual hemifields along the horizontal meridian.Stimulus presentation was equiprobable. ISI randomly var-ied among 1000 and 1500 ms. A white cross at the centerof the screen served as fixation point.

2.3. Procedure

Subjects were comfortably seated within a sound atten-uated, electrically shielded, and dimly illuminated cham-ber. The high resolution screen of a 486 IBM-compatiblecomputer was located outside the chamber in front of awindow, at the viewing distance of 114 cm from thesubject.

The task consisted in paying selectively attention and torespond by pressing a button as accurately and fast as

Žpossible to the S or H letter at a given level i.e., local or. Žglobal independent of visual field of presentation left or

.right . Due to the combination of letters at the two levels,

when attention was paid to either the large or small letters,stimuli were considered respectively global, or local ‘con-

Žgruent’ if they had the same identity at the two levels i.e.,.SrS or HrH , or ‘incongruent’ if they had different

Ž .identity at the two levels i.e., SrH and HrS . Theexperiment consisted in the presentation of 40 sequencesof 80 stimuli. At the beginning of each sequence, whichlasted about 2 min, subjects were instructed about the

Ž . Ž .target letter i.e., S or H and the level i.e., global or localthat they had to attend, as well as about the hand that theywere supposed to use. They were told to maintain theirgaze on the central fixation point and to minimize any kindof body or eye movements throughout each experimentalrun. Each sequence was followed by a short pause, whileevery 15 min a longer pause was given to the subjects.Stimuli, responding hand and attention conditions wererandomized across experimental runs and counterbalancedacross subjects.

2.4. ERP recording

Ž .The electroencephalogram EEG was continuouslyrecorded from 28 scalp sites using tin electrodes mounted

Ž .in an elastic cap Electro-cap . The electrodes were located

Table 1Summary of repeated measure ANOVAs and statistical significances

Data ANOVA Factors Significance p value

RTs 4 ways A, C, F, R A 0.0002C 0.0001A=C 0.001

Ž .N115 ampl. 5 ways A, C, F, E, H E 0.02A=C 0.04

Ž .N140 frontal lat. 5 ways A, C, F, E, H A 0.07C 0.06

Ž .N140 frontal ampl. 5 ways A, C, F, E, H E 0.03Ž .N180 lat. 5 ways A, C, F, E, H A 0.01

A=H 0.04Ž .N180 ampl. 5 ways A, C, F, E, H F=H 0.025

A=E 0.02Ž .P300 300–620 ms lat. 5 ways A, C, F, E, H A 0.003

E 0.01Ž .P300 300–620 ms ampl. 5 ways A, C, F, E, H A=E 0.05

Ž .Central–parietal P300 lat. 5 ways A, C, F, E, H A 0.003Ž .Occipital P300 lat. 4 ways A, C, F, H A 0.01

Ž .P3-I 300–460 ms ampl. 5 ways A, C, F, E, H A 0.025E 0.0002C=H 0.02

Ž .P3-II 460–620 ms ampl. 5 ways A, C, F, E, H A 0.008E 0.0001C=H 0.01A=E 0.03

( )Difference waÕesSN amplitude 4 ways A, C, E, H A=E 0.0007180–310 ms A=C=H 0.04SN latency 4 ways A, C, E, H A 0.05

E=H 0.03

AsAttention level: global, local; Cscongruence: congruent, incongruent; Fsvisual field: LVF, RVF; Rs responding hand: left and right; HsŽ .hemisphere: LH, RH; Eselectrode site: depending on the ERP component see text for specifications .


Fig. 1. Mean reaction times and standard deviations of responses toglobal and local targets as a function of congruence with the unattendedlevel.

Ž . Žat frontal Fp1, Fp2, FZ, F3, F4, F7, F8 , central CZ, C3,. Ž . ŽC4 , anterior-temporal T3, T4 , posterior-temporal T5,. Ž . ŽT6 , parietal PZ, P3, P4 , and occipital scalp sites OZ,

.O1, O2 of the International 10–20 System. Additionalelectrodes were placed half the distance between anterior-

Ž .temporal and central sites FTC1, FTC2 , central andŽ .parietal sites CP1, CP2 , anterior-temporal and parietal

Ž .sites TCP1, TCP2 , and posterior-temporal and occipitalŽ . Žsites OL, OR . Blinks and vertical eye movements i.e.,.EOG were monitored by means of two electrodes placed

below and above the right eye, while horizontal move-ments were recorded from electrodes placed at the outercanthi of the eyes. Linked earlobes served as referencelead. The EEG and the EOG were amplified with ahalf-amplitude band pass of 0.1–70 Hz and 0.01–70 Hz,respectively.

2.5. ERP data analysis

Continuous EEG and EOG were digitized for 100 msprior and 1000 ms following each stimulus presentation, ata rate of 512 samplesrs. They were stored with stimulusand response codes for off-line analysis. Before averaging,a computerized artefact rejection was performed to discardepochs in which eyes or muscle artefacts occurred. Aver-age ERPs were separately computed for each of the fourstimuli under the different attention conditions in the twovisual fields. The major ERP components were identifiedand measured automatically by a computer program, andquantified by measuring peak latency and baseline-to-peakamplitude and mean area values.

N115 component, prominent at occipital–temporal sites,was defined as the most negative peak between 80 and 140ms; N140 as the most negative peak between 120 and 180ms at frontal–central sites; N180 as the most negative peakbetween 145 and 215 ms at posterior occipital–temporalsites; P2 as the most positive peak between 180 and 260ms, and P3 as the most positive peak between 300 and 620ms. Mean area values were also measured for the two time

Žwindows in the P3 latency corresponding to P3-I 300–460. Ž .ms and P3-II 460–620 ms . Difference-waves were ob-

tained by subtracting nontarget from target responses bothfor congruent and incongruent configurations. Selection

Ž .negativity SN was measured as the mean amplitude areabetween 180 and 310 ms at temporal–occipital electrodesites.

RT and ERP measures associated with ocular and mus-cular artefacts, misses, and false alarms, were excludedfrom statistical analyses. For each subject, response timefaster than 140 ms or exceeding 2 standard deviations fromthe mean were also excluded from the analyses. Both RTand ERP measures were analyzed with repeated-measure

Ž .analyses of variance ANOVAs , adjusting for nonspheric-ity with the Greenhouse–Geisser epsilon coefficient. Fac-

Ž .tors were ‘attention level’ global and local , ‘congruence’Ž .congruent and incongruent , ‘visual hemifield’ of pre-

Ž . Žsentation left and right and ‘response hand’ left and.right . Additional factors were considered for ERP mea-

Žsures. They were ‘electrode site’ i.e., O1–O2, OL–OR,and T5–T6 for N115 and N180 posterior components;

Table 2Mean reaction times and standard deviations for global and local targets as a function of congruence with the unattended level, visual field, and respondinghand

Global Local

Congruent Incongruent Congruent Incongruent

RVF LVF RVF LVF RVF LVF RVF LVF

LH RH LH RH LH RH LH RH LH RH LH RH LH RH LH RH

380 383 382 385 381 385 385 388 415 422 422 428 455 466 472 47236 27 32 25 36 30 31 24 41 34 42 34 48 38 47 37


C3–C4, F3–F4, F7–F8 sites for anterior N140; T3–T4,OL–OR for SN; C3–C4, P3–P4, T3–T4, OL–OR, O1–O2

. Žfor the P3-I and P3-II components and ‘hemisphere’ left.and right . Student’s t-test were carried out for post-hoc

comparisons between means. Table 1 summarizes all thestatistical analyses with their significant factors.

3. Results

3.1. BehaÕioral results

RTs were significantly faster in the global than localŽ .condition see Fig. 1 as proven by the significant main

Ž . Ž .Fig. 2. Grand-averaged ERPs to global top and local bottom targets as a function of congruence with the unattended level. Recordings are from OZelectrode site.


Table 3Ž .Mean amplitude values mV and standard deviations of N115 compo-

nent for global and local targets as a function of congruence with theunattended level

Global Local

Congruent Incongruent Congruent Incongruent

y0.84 y0.96 y1.07 y0.752.0 2.0 2.2 2.1

Ž .‘attention level’ factor F 1,6s60.2; p-0.0002 . TheŽ .significance of ‘congruence’ F 1,6s72.6; p-0.0001

also indicated that RTs were faster to congruent thanincongruent stimuli. Furthermore, the ANOVA yielded asignificant interaction of ‘attention level’=‘congruence’Ž .F 1,6s33.15; p-0.001 . Post-hoc comparisons indi-cated that the incongruent level negatively affected speed

Ž .of response only in the local attention level p-0.01 ,thus showing an effect of interference of the global towardthe local level. No effect of visual field was found, al-though RTs to local targets tended to be faster whenstimuli were presented in the right than left visual hemi-

Ž .field see means of Table 2 .

3.2. Electrophysiological results

In Fig. 2 are displayed the grand-average ERPs toŽ . Ž .global top and local bottom targets as a function of

congruence. As can be clearly seen, the waveshape mor-phology at an early latency stage is characterized by aprominent negative deflection, peaking at about N115 msand followed by a tiny somewhat later positive deflectionpeaking at about 150 ms. This particular morphology

Ždiffers from the classical positive–negative–positive P–.N–P complex described in VEP literature, but may be

often found when using large bright patterns for stimula-Žtion, especially with a high spatial frequency content as in

w x.Zani and Proverbio’s VEP study 24 . Since this compo-nent represented the most significant source of potentialvariation in the early latency range, it was considered forstatistical analysis.

ANOVA performed on peak amplitude values of N115,most prominent at mesial-occipital, lateral-occipital andposterior-temporal sites, revealed a significant effect of

Ž .electrode site F 2,12s5.3; p-0.02 , indicating for thiscomponent the focus of maximum amplitude at mesial-oc-

Ž .cipital sites see Fig. 3 . Most interestingly, the analysis

Table 4Ž .Mean amplitude values mV and standard deviations of N180 compo-

nent recorded over the two cerebral hemispheres as a function of stimulusvisual hemifield

Left hem. Right hem.

LVF RVF LVF RVF

y3.8 y3.5 y4.5 y4.33.0 3.2 3.0 3.1

also yielded a significant interaction of ‘attention level’=Ž .‘congruence’ F 1,6s7.13; p-0.04 . Post-hoc t-tests

Ž .p-0.03 indicated that, when paying attention to thelocal level, sensory-evoked responses were significantlysmaller if the latter was part of an incongruent rather thancongruent configuration. Conversely, no difference was

Žfound for global patterns as a function of congruence see.Table 3 . These different effects of congruence as a func-

tion of the attended level can be clearly appreciated in Fig.2.

The negative component N180 focused over secondaryŽ .visual areas OL, OR; T5, T6 , showed an earlier peak in

Žthe global than local condition F 1,6s13.4, p-0.01;.globals185"16, locals192"16 ms . This peak was

Žlarger to global than local stimuli at lateral-occipital p-. Ž .0.01 , and posterior-temporal sites p-0.02 , but not at

Žmesial-occipital sites ‘attention level’=‘electrode’; F.2,12s5.4, p-0.02 . Furthermore, it was of greater am-

Ž .plitude over the right than left hemisphere p-0.0001Ž .independent of visual field of presentation see Table 4 , as

Ž .indicated by the post-hoc comparisons p-0.025 for theŽsignificant interaction of ‘hemisphere’=‘visual field’ F

.1,6s8.8; p-0.025 .Over fronto-central sites the earlier negativity N140

showed a strong tendency to be faster to congruent thanŽ .incongruent configurations F 1,6s5.04; ps0.06 , and

Ž .to global than local ones F 1,6s4.7; ps0.07 . Table 5shows mean N140 amplitudes and standard deviations forthese conditions.

P300 component, broadly distributed over central–parietal areas, was extremely affected both in latency andin amplitude by ‘attention level’ and ‘congruence’. P300 toconfigurations attended at the global level were muchfaster and larger than those to configurations attended at

Ž .the local level see Fig. 4 . P3 peaked earlier for globalŽthan local targets both at parietal–central F 1,6s22;

. Žp-0.003 and occipital scalp sites F 1,6s12.7; p-

Fig. 3. Grand-average isocontour voltage maps of early sensory response to local targets in the latency range of N115 component. Note that theincongruence of the local level with the global one significantly affected the sensory-evoked activity of primary visual areas. The color scale indexes thepolarity and amplitudes of the topographical maps.

Fig. 4. Grand-averaged ERPs to global and local targets as a function of recording site. As a whole, P300 was much earlier and larger to global than localtargets.


Table 5Ž .Mean latency values ms and standard deviations of anterior N140

component as a function of target congruence and visual field

Congruent Incongruent Global Local

138 141 137 14217 17 17 17

.0.01 , as indicated by statistical analysis performed onŽ .peak latency see means of Table 6 . Furthermore, P300

Žwas faster at parietal than central sites F 1,6s13.6;.p-0.01 . The ANOVA performed on peak amplitude

showed a significant interaction between ‘attention level’Ž .and ‘electrode site’ F 1,6s6; p-0.05 . Post-hoc com-

parisons between means indicated that P300 differed in its

Ž . Ž .Fig. 5. a Grand-average ERPs to global and local targets as a function of stimulus congruence as recorded at the F8 right lateral frontal site. bGrand-average ERPs to global and local targets as a function of stimulus congruence as recorded at the F7 left lateral frontal site. Note that the largepositive complex has two phases in the local attended condition.


Table 6Ž .Mean latency values ms and standard deviations of P300 component as

a function of recording site and attended level

Parieto–central Occipital

Global Local Global Local

398 446 378 43051 71 48 68

distribution as a function of the attended level, it beingŽ .more anteriorly distributed for the local one p-0.01 .

It is interesting that this broad positivity showed twophases, the first being larger to the global targets, and thesecond to the local ones. In addition, they showed differenttopographical distributions, and were clearly differentiatedin two sub-components in many of the subjects. For thesereasons, separate ANOVAs were performed on mean areavalues computed within two different time-windows: from

Ž . Ž300 to 460 ms i.e., P3-I , and from 460 to 620 ms i.e.,.P3-II .

Attention to local elements gave rise to a larger P3-IIŽ .‘attention level’ factor: F 1,6s15; p-0.008 , while

Žattention to global patterns gave rise to a larger P3-I F.1,6s8.5; p-0.025 . For both time windows, P3 showed

larger amplitudes to congruent than incongruent patterns,but a different hemispheric lateralization. Indeed, the sig-

Table 7Ž .Mean latency values ms and standard deviations of selection negativity

as a function of cerebral hemisphere and recording site

Left hem. Right hem.

Temporal Occipital Temporal Occipital

269 250 259 26072 69 77 81

Žnificant interactions of ‘congruence’=‘hemisphere’ F.1,6s8.9; p-0.02 indicated that P3-I was larger to

congruent than incongruent patterns over the right hemi-Ž .sphere p-0.0001 , whereas P3-II showed the same pat-

Žtern of results over the left hemisphere F 1,6s12.3;.p-0.01; post-hoc test: p-0.006 . These asymmetric dis-

tributions can be seen comparing the right-sided wave-forms of Fig. 5a with the left-sided ones of Fig. 5b. Mapsof Fig. 6 show the spatio-temporal progression of brainactivation linked to target congruence with the unattendedlevel. Global congruent targets elicited a significantly en-hanced response in the P3-I phase especially over right

Žcentro–parietal sites see for instance the upper row around.380 ms , whereas local congruent targets displayed an

enhanced response at left frontal sites in the P3-II phaseŽ .see for instance the lower row around 560 ms . Thus, theeffect of congruence in P300 showed a posterior distribu-

Ž .Fig. 6. Isocontour voltage maps of brain activation in the P300 latency range 300–620 ms . Maps were computed on difference waves obtained bysubtracting ERPs to incongruent from those to congruent targets. The color scale indexes the polarity and amplitudes of the topographical maps. Note that

Ž . Ž .P3-I 300–460 ms was more sensitive to congruence over the right centro–parietal sites for global targets, while P3-II 460–620 ms was more sensitiveto congruence over anterior left sites.


Ž . Ž .Fig. 9. Isocontour voltage maps of difference waves for the global upper maps and local lower maps congruent conditions. Difference waves wereobtained by subtracting ERPs to nontargets from ERPs to targets. The color scale indexes the polarity and amplitudes of the topographical maps. Upperand lower left maps show brain activation relative to selection negativity, whereas upper and lower right maps show brain activation relative to P300component. At all latency ranges the maps indicate clearcut spatio-temporal differences in the progression of brain activation dependent on the attentioncondition. Selection negativity was much earlier to global than local targets, and had a right anterior temporal distribution for the former ones, and a leftoccipital–temporal distribution for the latter ones. As for P300 component, in the upper map it is visible P3-I to global targets focused at posterior sites,with an evident asymmetry over the right hemisphere. The lower map shows P3-II to local targets strongly lateralized over left centro–frontal sites.

Ž .tion in the first phase P3-I for global targets, and anŽ .anterior distribution in the later phase P3-II for local

targets.

3.3. Difference waÕes

The attention effects per se appeared after about 140 mspost-stimulus latency as a negative potential having aposterior distribution and being followed by a later positiv-

ity. Difference waves, obtained by subtracting ERPs tonontargets from those to targets, evidenced a posterior

Žnegativity specific to target selection selection negativity.sSN , peaking on average at about 259 ms. ANOVA

performed on peak amplitude revealed that SN was muchŽ . Ž .faster to global 227"56 ms than local 292"76 ms

Ž .targets F 1,6s5.5; p-0.05 . In addition, it was earlierŽover occipital than temporal left sites post-hoc test: p-

.0.01 , whereas it was about of the same latency over the

Fig. 7. Difference waves for global congruent stimuli obtained by subtracting ERPs to nontargets from those to targets. Overlapped are waveformsŽ . Ž .recorded from middle-temporal i.e., T3, T4 and lateral-occipital i.e., OL, OR electrode sites.

Fig. 8. Difference waves for local congruent stimuli obtained by subtracting ERPs to nontargets from those to targets. Overlapped are waveforms recordedŽ . Ž .from middle-temporal i.e., T3, T4 and lateral-occipital i.e., OL, OR electrode sites.


right hemisphere, as shown by the significant interaction ofŽElectrode=Hemisphere F 1,6s7.6; p-0.03; see means

.of Table 7 .Interestingly, ANOVA performed on mean area ampli-

tude of this component supported the evidence of ananatomical specificity, in that SN reached its maximum

Žamplitude at middle-temporal sites for global targets post-.hoc test: p-0.01 , and at lateral-occipital sites for local

Ž .targets p-0.05 , as indicated by the significant interac-Žtion of ‘electrode’=‘attention level’ F 1,6s41; p-

.0.0007 . Furthermore, its amplitude changed significantlyas a function of the interaction between task, congruence,

Ž .and hemisphere F 1,6s6.4; p-0.04 . Indeed, for con-gruent patterns, target selection resulted in an increased SNasymmetrically distributed over the right hemisphere when

Žattention was paid to the global patterns p-0.01; see.Fig. 7 , and over the left hemisphere when attention was

Ž .paid to the local elements p-0.05; see Fig. 8 . Forincongruent patterns, SN was also larger over the right

Ž .hemisphere for global targets post-hoc test: p-0.05 , butno asymmetry was evident for local incongruent targets.Further post-hoc comparisons indicated that incongruenceaffected only targets at the local level over the left hemi-sphere, with congruent targets leading to a bigger SN than

Ž .incongruent ones p-0.01 .Basically, then, this component showed a very different

time-course and topography as a function of the attendedlevel. It was much earlier for global targets and reached itsmaximum amplitude at the right temporal site, whereas,during selection of the local level, it was later and dis-tributed more posteriorly at occipital sites, with a left-sidedasymmetry for congruent targets.

Maps of Fig. 9 summarize the effects of target selectionconveyed by target–nontarget difference waves, by show-ing the spatio-temporal distribution of selection negativityand P300 to global and local congruent targets. It isevident the anterior distribution of attention effects for the

Ž .late positive phase of P300 component P3-II .

4. Discussion

Overall, the behavioral data obtained in the presentstudy showed a strong advantage of global over localtargets, along with an interference effect for local, but notglobal, incongruent targets. The electrophysiological re-sults were consistent with such a pattern of results. Inaddition, they provided robust evidence for the hypothesisof a sensory advantage of information conveyed by theglobal vs. local level. They also indicated an early modula-tion of brain electrical activity of visual areas during anattention task based on features discrimination. These ef-fects are consistent with early latency attention effectsshown in literature for selection of check size and spatial

w xfrequency gratings 24,25 as well as alphanumeric andw xgeometrical stimuli 23 . In the present experiment, early

N115 responses linked to primary visual areas activationshowed to be differentially modulated by stimulus incon-gruence and attended level. The diminished visualevoked-response to local targets incongruent with the globalconfiguration indicates a dominance of global spatial infor-

w xmation, as suggested by Hughes et al. 9 . The electrophys-iological techniques do not allow us to ascertain definitelywhether this advantage is due to the faster conductiontimes and greater sensitivity gain of low frequency chan-nels, or to other factors, but certainly the data suggest thatthe information conveyed by the latter channels is capableof inhibiting sensory processing of local elements.

This asymmetry in visual processing for inputs con-veyed at the two perceptual levels was evident also at laterprocessing stages as indexed by N1, SN and P3 compo-nents. Occipital N180 was significantly faster and larger toglobal targets, whereas P300 was much earlier to global

Ž .targets and larger to them in its first phase P3-I . Interest-ingly, for both components it appeared an anatomicaldissociation suggesting the existence of two separate pro-cessing channels. N180 was larger to global than localtargets at lateral occipital and infero-temporal sites, whereasthis difference disappeared at mesial-occipital sites. Thelarge positive P3-I had a more posterior distribution com-pared to P3-II. Again, the effect of congruence with theunattended level, and the effect of attentional selectionŽ .target vs. nontarget showed an early posterior distributionfor P3-I and a later anterior distribution for P3-II phase ofP3 complex. This anterior–posterior distribution for P300component does not overlap with the classical distributionof early anterior P3a and posterior P3b elicited by novelirrelevant and target stimuli, respectively, as shown for

w x.example by Knight 11 . Indeed, P3-I and P3-II were bothelicited by target stimuli, and their differential distributionprobably reflects different cortical attentional selection sys-tems.

The finding of a different time course and topographyfor the SN provides further evidence for different selectionmechanisms for local vs. global aspects of visual informa-tion, in agreement with previous electrophysiological dataw x10 . Furthermore, this negativity related to target selectionshowed a task-dependent asymmetric lateralization. It wasfocused at temporal sites when attention was paid to theglobal level, and at occipital sites when attention was paidto the local level. Results showed that, when the unat-tended level did not interfere with the attended one, in thatstimuli were congruent at the two levels, the selection oflocal elements resulted in a strong left-sided occipitalactivation, whereas the selection of the global configura-tion resulted in a right-sided temporal activation.

These results are strongly consistent with neuropsycho-logical data providing evidence of a hemispheric special-ization for processing of localrglobal configurations. Forexample, a well-known series of neuropsychological stud-

w xies on unilateral brain damaged patients 14,21,22 hasshown that patients suffering from temporo–parietal le-


sions are impaired in global processing when the righthemisphere is damaged, and in local processing when theleft hemisphere was damaged. Moreover, the recent neu-

Ž . w xrometabolic study rCFB by Fink et al. 2 has clearlydemonstrated the role of temporal areas in the attentionalcontrol of global and local processing. In particular, targetselection at the global level resulted in an increased activ-ity of the right temporal–parietal–occipital junction,whereas target selection at the local level resulted in anincreased activity of the left posterior aspect of the supe-rior temporal gyrus. In addition, the authors found that in ablocked condition, where the target level was kept fixedthroughout the run, there was an increased metabolic activ-ity at the right lingual gyrus when attention was paid to theglobal level, and at the left inferior occipital cortex, whenattention was maintained to the local level. Thus, thedifferent involvement of left and right portions of Brod-mann’s area 18 in this rCBF experiment might indicatethat the hemispheric asymmetry is not confined to atten-tional selection mechanisms at higher cognitive levels, butis present at perceptual level too.

The electrophysiological results obtained in the presentstudy are somewhat different from those of Heinze and

w x w xMunte 3 and Johannes et al. 10 , who failed to find a¨clear global advantage effect at either behavioral or elec-trophysiological level. It has to be mentioned that thesestudies adopted divided attention tasks in which subjectswere forced to switch their attention from one level to theother one on a trial-by-trial basis, because targets could be

w xat either levels. With such a paradigm, Johannes et al. 10found RTs of the same speed for targets of both levels andno interference effects. This was true even for brain poten-

w xtials. Heinze and Munte 3 , who adopted hierarchical¨stimuli of different sizes, found a size-dependent prece-

Ždence effect i.e., a local advantage for 78, and a global.advantage for 28 stimuli , but no clear interference effects.

The onset of a selection-related N250 was earlier for localthan global targets in the 78 condition, but equal for thetwo levels in the 28 condition. Overall, it seems thatswitching the attentional focus from one level to the otherone might have reduced the effect of global precedenceand a clear pattern of asymmetric interference. At these

w xregards, it is important to mention that Robertson 20 , in aseries of behavioral experiments with hierarchical stimuli,clearly demonstrated the effect of attentional persistenceŽ .i.e., level-specific priming in globalrlocal selection. Theauthor found that switching from a larger global target to asmaller local target or vice-versa slowed response timecompared to conditions where the target level remained thesame. Very interestingly, it was shown that this primingeffect was related to the stimulus spatial frequency content.Indeed, the specific sequential effects due to the attentionalpersistence were eliminated by filtering the low spatialfrequencies in the pattern preceding a not-filtered hierar-

w xchical pattern. Similarly, Hubner 6 found a clear effect of¨stimulus spatial frequency in interaction with task switch-

ing. When the target level was kept constant throughout aŽ .run e.g., Expt. 2 , he found a normal advantage of the

global configuration, which disappeared in the case ofhierarchical high-passed filtered stimuli. The same did notoccur when the target level was made to vary on a

Ž .trial-by-trial basis e.g., Expt. 1 . Then, it seems asglobalrlocal processing is greatly affected by both sensoryand attentional variables. It is very interesting, at this

w xregard, that rCBF data by Fink et al. 2 have shown thatdifferent brain areas are involved in globalrlocal process-ing during fixed vs. switching attention tasks.

Overall, the data of the present study suggest the inter-vention of late-latency attentional factors in determininghemispheric asymmetries for the two processing levels.Indeed, a clear pattern of hemispheric dominance appearedfor the temporal–occipital SN and later central–parietalP300 component. In this regard, a large behavioral litera-ture supports the view that a hemispheric asymmetry mayonly arise in tasks where spatial frequency processing is

Žactually demanded see for example the review by Helligew x w x w x.5 , Navon 16 and Proverbio et al. 17 , which meansthat a frequency-based selection is required by the task.However, there is also electrophysiological evidence ofpure sensory-perceptual based hemispheric asymmetries in

w xthe processing of low vs. high frequency stimuli 18,24 .Indeed, the occipital locus for brain activation during

w x Žglobal vs. local selection reported by Fink et al. 2 i.e.,.right lingual gyrus and left inferior occipital cortex seems

to support the involvement of perceptual mechanisms inthe hemispheric specialization. However, further investiga-tion is needed to reach a definitive conclusion on thisissue.

In conclusion, the current study provided an instance ofperceptual dominance of global level of visual information.This is supported by the finding of an advantage of globaltargets both in terms of speed of RTs, and latency andamplitude of ERP components. Interestingly, electrophysi-ological results provided evidence of an early interferenceeffect, as indexed by the modulation of a N115 compo-nent, of the global incongruent information over the localone. Last but not least, evidence was also provided for twoseparate processing channels, subserved by distinctanatomical systems asymmetrically lateralized, for the at-tentional selection of global vs. local elements of hierarchi-cally organized visual patterns.

Acknowledgements

The present work was carried out at the Department ofPsychology of University of Trieste, and funded byMURST grants. It was also supported by CNR grants to A.Zani. We wish to thank Alessandro Vegliach and CarmelaAvella for technical assistance.


References

w x1 C. Deruelle, H.J. Neville, Different neural systems are active duringglobal and local visual processing: a RT and ERP study, Abstr.

Ž .Book Cogn. Neurosci. Soc. 1994 53.w x2 G.R. Fink, P.H. Halligan, J.C. Marshall, C.D. Frith, R.S.J. Frack-

owiak, R.J. Dolan, Where in the brain does visual attention selectŽ .the forest and the trees?, Nature 382 1996 626–628.

w x3 H. Heinze, T.F. Munte, Electrophysiological correlates of hierarchi-¨cal stimulus processing: dissociation between onset and later stages

Ž .of global and local target processing, Neuropsychologia 31 1993841–852.

w x4 H. Heinze, S. Johannes, T.F. Munte, G.R. Mangun, The order of¨global and local-level information processing: electrophysiologicalevidence for parallel perceptual processes, in: H. Heinze, T. Munte,¨

Ž .G.R. Mangun Eds. , Cogn. Electrophysiology, Birkhauser, Boston,1994, pp. 102–123.

w x5 J.B. Hellige, Hemispheric asymmetry for components of visualŽ .information processing, in: R.J. Davidson, K. Hugdahl Eds. , Brain

Asymmetry, The MIT Press, Cambridge, MA, 1995, pp. 99–122.w x6 R. Hubner, The effect of spatial frequency on global precedence and¨

Ž .hemispheric differences, Percept. Psychophys. 59 1997 187–201.w x7 H.C. Hughes, Asymmetric interference between components of

Ž .suprathreshold compound gratings, Percept. Psychophys. 40 1986241–250.

w x8 H.C. Hughes, R. Fendrich, P.A. Reuter-Lorentz, Global versus localprocessing in the absence of low spatial frequencies, J. Cogn.

Ž .Neurosci. 2 1990 272–282.w x9 H.C. Hughes, G. Nozawa, F. Kitterle, Global precedence, spatial

frequency channels and the statistics of natural images, J. Cogn.Ž .Neurosci. 8 1996 197–230.

w x10 S. Johannes, B.M. Wieringa, M. Matzk, T.F. Munte, Hierarchical¨visual stimuli: electrophysiological evidence for separate left hemi-spheric global and local processing mechanisms in humans, Neu-

Ž .rosci. Lett. 210 1996 111–114.w x11 R.T. Knight, Distributed cortical network for visual attention, J.

Ž .Cogn. Neurosci. 9 1997 75–91.w x12 L.L. Lagasse, Effects of good form and spatial frequency on global

Ž .precedence, Percept. Psychophys. 53 1993 89–105.

w x13 M.R. Lamb, L.C. Robertson, Do response time advantage andinterference reflect the order of processing of global- and local-level

Ž .of information?, Percept. Psychophys. 46 1989 254–258.w x14 M.R. Lamb, L.C. Robertson, R.T. Knight, Component mechanisms

underlying the processing of hierarchically organized patterns: infer-ences from patients with unilateral cortical lesions, J. Exp. Psychol.

Ž .Learn. Mem. Cogn. 16 1989 471–483.w x15 W.J. Lovegrove, S. Lehmkuhle, J.A. Baro, R. Garzia, The effects of

uniform field flicker and blurring on the global precedence effect,Ž .Bull. Psychon. Soc. 29 1991 289–291.

w x16 D. Navon, Forest before trees: the precedence of global features inŽ .visual perception, Cogn. Psychol. 9 1977 353–383.

w x17 A.M. Proverbio, A. Zani, C. Avella, Hemispheric asymmetries forspatial frequency discrimination in a selective attention task, BrainCogn., 1997.

w x18 A.M. Proverbio, A. Zani, C. Avella, Differential activation of multi-ple current sources of foveal VEPs as a function of spatial fre-

Ž .quency, Brain Topogr. 9 1996 59–68.w x19 K.R. Ridderinkhof, M.W. van der Molen, When global information

and local information collide: a brain potential analysis of the locusŽ .of interference effects, Biol. Psychol. 41 1995 29–53.

w x20 L.C. Robertson, Attentional persistence for features of hierarchicalŽ .patterns, J. Exp. Psychol. Gen. 125 1996 227–249.

w x21 L.C. Robertson, M.R. Lamb, R.T. Knight, Effects of lesions oftemporo–parietal junction on perceptual and attentional processing

Ž .in humans, J. Neurosci. 8 1990 3757–3769.w x22 L.C. Robertson, M.R. Lamb, Neuropsychological perspectives on

Ž .theories of partrwhole organization, Cogn. Psychol. 23 1991299–330.

w x23 W. Skrandies, Information processing and evoked potentials: topog-Ž .raphy of early and late components, Adv. Biol. Psychiat. 13 1983

1–12.w x24 A. Zani, A.M. Proverbio, ERP signs of early selective attention

effects to check size, Electroencephalogr. Clin. Neurophysiol. 95Ž .1995 277–292.

w x25 A. Zani, A.M. Proverbio, Attention modulation of short latencyERPs by selective attention to conjunction of spatial frequency and

Ž .location, J. Psychophysiol. 11 1997 21–32.

research report - wordpress.com€¦ · cognitive brain research 6 1998 321–334 . research report...

Documents